Deep Neural Network Pruning for Nuclei Instance Segmentation in Hematoxylin & Eosin-Stained Histological Images

TL;DR

This study explores the effects of magnitude-based pruning techniques on deep neural networks used for nuclei instance segmentation in histological images, demonstrating significant weight reduction with minimal performance loss.

Contribution

It is the first to evaluate layer-wise and network-wide pruning impacts on a specialized medical image segmentation model, revealing their effectiveness and trade-offs.

Findings

Layer-wise pruning performs slightly better at small compression ratios.

Network-wide pruning outperforms at large compression ratios.

Up to 93.75% of weights can be pruned with less than 2% performance loss.

Abstract

Recently, pruning deep neural networks (DNNs) has received a lot of attention for improving accuracy and generalization power, reducing network size, and increasing inference speed on specialized hardwares. Although pruning was mainly tested on computer vision tasks, its application in the context of medical image analysis has hardly been explored. This work investigates the impact of well-known pruning techniques, namely layer-wise and network-wide magnitude pruning, on the nuclei instance segmentation performance in histological images. Our utilized instance segmentation model consists of two main branches: (1) a semantic segmentation branch, and (2) a deep regression branch. We investigate the impact of weight pruning on the performance of both branches separately and on the final nuclei instance segmentation result. Evaluated on two publicly available datasets, our results show that layer-wise pruning delivers slightly better performance than networkwide pruning for small compression ratios (CRs) while for large CRs, network-wide pruning yields superior performance. For semantic segmentation, deep regression and final instance segmentation, 93.75 %, 95 %, and 80 % of the model weights can be pruned by layer-wise pruning with less than 2 % reduction in the performance of respective models.